What rights does a consumer have to a digital product that they purchase? What control does an author have over their creative works? What obligations do others have to you when they use your data to develop new products?
None of these questions are new, but the recent proliferation of artificial intelligence (AI) tools, particularly generative AI (GenAI) tools, has brought these questions to the forefront of ongoing conversations about the role that AI will play in the marketplace. As an advocate for and enforcer of consumer protection laws, the US Federal Trade Commission (FTC) recently offered reminders and recommendations for companies offering digital products and AI tools:
Manage Consumer Expectations About Their Rights to Digital Products
Upon purchasing a product, a consumer may have certain expectations about their rights to that product. However, sellers, particularly of digital products, may limit those rights in ways not readily apparent to the consumer. Such limitations could include selling the product only under a limited, revocable license, limiting the right to repair the product, retaining the ability to control software updates or functionality of the product, providing access to product features only via subscription, changing terms of use after purchase, or restricting access to the product after death of the consumer. Regardless of the types of restrictions sellers place on their products, the FTC recommends making those restrictions clear to consumers prior to purchase. Failure to be up front about any restrictions could make a seller liable for deceptive practices in violation of the FTC Act.
Avoid Deceptive Practices With Respect to Creative Works
Similarly, authors may have certain expectations about their rights with respect to their creative works. For example, they may expect that others should not be permitted to pass off their work as created by that author. As a matter of consumer protection, the FTC seeks to prevent others from committing such deception. But GenAI tools make it both easier and faster to create works that are harder to distinguish from the works of another author. Try using one of the GenAI tools currently available to write a haiku in the voice of Langston Hughes or reproduce a photograph as a painting in the style of Frida Kahlo and you may be surprised how convincing the tool is at mimicking Hughes’ voice or Kahlo’s style.
Authors may also expect to retain full ownership of creative works that they publish online. Whether that is true depends on the applicable publishing site’s terms of use. Failure to make clear and to honor the authors’ rights to their published content under the terms of use may be considered a deceptive practice violative of the FTC Act.
Consider Being Transparent About Training Sources for GenAI Tools
GenAI tools are developed by using data to train an algorithm. The content and quality of the data influences the effectiveness of the tool. In their quest to develop GenAI tools able to perform a variety of tasks, developers pull from a variety of data sources, which may include creative works such as poems, movie scripts, blog posts, and even news articles, as well as personal information protected by privacy laws. From a consumer protection perspective, the FTC is concerned that lack of transparency about training data sources may be deceptive because individual and corporate users are not able to fully evaluate whether a given GenAI tool is suitable for their needs.
Conclusion
While the rise of AI has resurfaced old questions about ownership, control, and product development, the FTC’s reminders and recommendations to avoid deception by clearly and accurately communicating with consumers and creators are evergreen. Baker McKenzie’s AI practice stands ready to assist companies as they navigate these old issues arising from new technologies.
